Prospective association of tobacco retailer density and neighbourhood walkability with youth vaping initiation in California, USA

Abstract

Background: Tobacco retailer density might influence youth e-cigarette use due to increased access and exposure to point-of-sale marketing. There is a need for longitudinal investigations on the association of tobacco retailer density with youth e-cigarette use, with consideration of contextual factors such as neighbourhood walkability that could enhance retailer exposure.

Methods: Five semi-annual waves (Fall 2021-Fall 2023) of a Southern California school-based cohort of youth who never vaped at baseline (n=3401; mean baseline age=15 years [range=12-17]) were merged with spatial data on tobacco retailers corresponding to each school year. Time-varying measures of retailer exposures included count of retailers within circular home buffers (800 m/1600 m buffers) and retailer density/km² within home census tracts. Discrete-time survival analyses estimated associations of time-lagged and time-varying retailer exposures with vaping initiation at each 6-month follow-up, adjusting for race/ethnicity, parental education, neighbourhood walkability and an Area Deprivation Index. Secondary analyses examined moderation by neighbourhood walkability based on a National Walkability Index.

Results: The incident rate of vaping initiation was 3.8% per 6-month follow-up. After covariate adjustment, only associations with census tract retail density were statistically significant (risk ratio (RR) highest vs lowest quartile=1.36, 95% CI=1.01 to 1.84). In moderation analyses, count of retailers within 800 m home buffers was positively associated with vaping initiation for youth who lived in the most walkable neighbourhoods (RR=1.08, 95% CI: 1.01 to 1.16), with no association for youth in less walkable neighbourhoods (RR=1.00, 95% CI:0.95 to 1.05; interaction p-value=0.0238).

Conclusion: Findings highlight the potential role of restricting retailer density in preventing youth vaping, particularly in areas characterised by both greater walkability and tobacco retailer density.

Keywords: Electronic nicotine delivery devices; Environment; Nicotine; Non-cigarette tobacco products.

Figure 1.. Flowchart of exclusions, censoring, and e-cigarette initiation over follow-up

Figure 2.. Joint association of tobacco retailer exposure and neighborhood walkability with e-cigarette initiation

Forest plot of results from models examining the joint association of tobacco retailer exposure (above vs. below median) and neighborhood walkability (most vs. less walkable) with e-cigarette initiation. Models are adjusted for exposure wave, parental education, race/ethnicity, and Area Deprivation Index deciles. Most walkable defined as ≥75^th percentile of national walkability index. Less walkable defined as <75^th percentile of national walkability index. See Supplemental Table 6 for corresponding statistics. *Indicates estimates are significantly different from one another (interaction p-value=0.0151); above median vs. below median among most walkable neighborhoods RR=1.70 (95%CI: 1.14–2.53)